Introduction: Microstructure control in metal additive manufacturing is ideal for achieving superior customized mechanical properties. Columnar to equiaxial transition engineering in the additive manufacturing process is critical to its technological advancement in the rapid solidification process. Here, we report a powder size-driven approach to melt pool engineering that demonstrates convenient and large-scale control of grain morphology by triggering a counterintuitive response of powder size to the microstructure of additively manufactured 316 L stainless steel. We obtained coarse-grained (>100 μm) or near-single-grain microstructures from fine powders, and near-isotropic fine-grained (<10 μm) microstructures from coarse powders. This approach shows a resourceful adaptability to directed energy deposition and powder bed fusion without increasing cost, where particle size-dependent powder stream preheating effects and powder bed thermophysical properties drive changes in microstructure. This work provides a pathway to achieve more controllable, economical, and sustainable manufacturing of metal additives using raw material particle size distribution.
Producing metal parts that outperform traditional parts has become a major research focus for metal additive manufacturing (AM). The mechanical properties of additively manufactured parts without machining defects are highly dependent on their microstructure. As a result, the ability to control the evolution of microstructures during part manufacturing facilitates the production of complex shapes with predictable mechanical properties. In addition, the ability to tailor microstructures should be adaptable to a wide range of AM technologies. The demands on stainless steel 316 L (SS316L) are particularly high due to their wide range of general use in many industrial sectors in harsh and corrosive environments1,2.
In this work, we explored the effect of PSD on the microstructural evolution of SS316L printed using the L-DED process by systematically varying the powder size under a constant set of processing parameters. This controlled study from monorail deposition to bulk samples helps to isolate the response of microstructural evolution to changes in PSD. Taking advantage of this role, we report a melt pool engineering (MPE) approach in which we demonstrate site-specific microstructure control using L-DED. In addition, we have achieved bidirectional control of the microstructure evolution in the E-PBF process by taking advantage of the effect of PSD changes on the thermophysical properties of the powder bed and thus on the solidification behavior of the melt pool. This approach enables coarse and fine PSDs to form fine isometric and coarse columnar microstructures, respectively, without increasing costs.
L-DED 工艺中原料粒度与预热温度的相关性：
In the L-DED process, simple and precise microstructure control is achieved within the same layer of the POC section. We show that the PSD of the feedstock affects the melt pool structure and thus the microstructure of the grain, as the particle temperature increases as it passes through the laser irradiated region and hits the melt pool.
PSD-driven microstructural changes applied to L-DED stainless steel samples:
We confirm that the microstructural responses of particle size in the L-DED monorail are opposite. To validate the powder size-driven MPE method, we made L-DED bulk samples using a mixture (FC) of fine, fine, and coarse powder (1:1 mix ratio) and coarse powder (Method and Supplementary Figure 2b). We found that both the depth and width of the melt pool measured in the transverse plane increased with the average particle size (Figure 3a-c). The melt pool depth measurements for fine, FC, and coarse samples were 380, 407, and 458 μm, respectively. Similarly, the melt pool widths are 657, 675, and 712 microns, respectively. A large amount of epitaxial grain growth was found in the fine powder sample (Figure 3a), and the well-oriented grains became coarse.
PSD-Driven Microstructure Control in PBF – An Innovation of the Traditional Approach to CET in Metal AM:
We envision that PSD-driven microstructural changes will effectively produce microstructural differences in cases where particle size may affect the overall cooling process, such as the PBF AM process. Considering the limitations of commercial L-PBF systems in the 15-63 μm particle size range, we chose the E-PBF process to establish our method's facilities for the production of fine equisquare microstructures, which have been widely used to produce near-single crystal microstructures in SS316L32 and nickel-based superalloys33,34. While our analytical calculations and numerical simulations support the intuitiveness of applying the L-DED method to site-specific microstructure control, it is not so intuitive to apply it to the E-PBF process due to the complexity of the effect of PSD on the thermophysical properties of the powder bed. To take the technology to the next level, we have developed a machine learning (ML) framework to study the effect of PSD on the thermophysical properties of powder beds with the aid of particle microstructure modeling.
Machine Learning Framework:
We found that there is a research gap regarding the effects of PSD on the microstructure and mechanical properties of PBF systems. This may be due to the lack of a framework that can predict the thermophysical properties of the powder bed, specifically the powder bed density (PBD) and thermal conductivity, based on the raw material PSD. To achieve this, machine learning (ML) was used to numerically simulate the heat transfer of a representative particulate microstructure of the sintered powder bed.
Effect of raw material PSD on the thermophysical properties of powder bed in E-PBF SS316L:
The bulk density of the coarse powder bed was 81%, which was higher than the 77% of the fine powder bed quantified by μ-CT measurement (Figure 4c and d). Thermal diffusivity values of 0.74 and 0.87 mm2 s-1 at 850 oC measured by laser flash spectrometer (LFA) were (Method and Supplementary Note 1). At 850 °C, the assay for undisturbed bulk samples was also ~6 mm2 s-1. At 450 °C, the specific heat capacity of the fine and coarse bed samples was also obtained with a specific heat capacity value of ~565 J kg-1K-1 independent of particle size. The temperature-dependent data and fitting curves obtained from these measurements are shown in Supplementary Figure 3.
Engineering CET in E-PBF 316 L stainless steel sample:
Based on our earlier study of E-PBF SS316 L8,32, we found that finely powdered raw materials printed with fully dense samples at optimal process parameters produced a coarse column microstructure and showed significant mechanical anisotropy. With our powder size-driven MPE method (the method), significant CETs were achieved from fine to coarse samples, as is evident from their respective EBSD micrographs (Fig. 5a, b). The microstructure of the fine powder E-PBF sample is close to that of single crystal coarse-grained, and the melt pool profile in the transverse plane is wide and shallow (w/d: 5.6). In stark contrast, a diffuse microstructure consisting of a large number of fine grains appears in a coarse powder sample with a semicircular melt pool profile (w/d:2) in the transverse plane.
Solidification pattern of powder size-driven MPE in L-DED and E-PBF processes:
We have demonstrated that, as we hypothesized, the powder size (or PSD) influences the evolution of the microstructure in both the L-DED and E-PBF processes, but more results are obtained in the E-PBF process due to the important role of the powder bed in the solidification of the melt pool and the creative E-PBF manufacturing strategy (method) we have adopted for both powders. In addition, examining the solidification conditions of fine and coarse powder samples printed with L-DED and E-PBF from the perspective of CET curves10,36,37 is key to revealing the underlying mechanisms of our method.
Adjusting the mechanical response of different microstructures by the powder-size-driven MPE method:
Representative engineering stress and strain curves obtained from uniaxial tensile tests on SS316 L samples made of E-PBF and L-DED (Fig. 6a) highlight the different mechanical responses of the coarse and fine structures obtained with fine and coarse powder feedstocks, respectively. Importantly, they highlight that if an unprecedented mechanical response is desired, it is essential to obtain as fine microstructures as possible, especially in the case of 3D printed stainless steel. Coarse-grained raw materials give both E-PBF and L-DED better performance.
由Shubham Chandra, Su Beng Tor等学者联合完成.
相共研究成果以“Powder-size driven facile microstructure control in powder-fusion metal additive manufacturing processes”发表在nature communications上
Link: https://www.nature.com/articles/s41467-024-47257-w#Sec24
Figure 1: PSD-driven grain morphology and size control in a 3D printed SS316L microstructure.
a Particle preheating temperature varies with particle size and laser power. b Schematic diagram showing flight time and temperature as a function of particle size. The color of the particles corresponds to temperature (brown to pink on the Batlow color palette 52 indicates the change from low to high temperatures, respectively). c Particle size preheating temperature change obtained at 300 W laser power, and PSD for fine and coarse powder measurements. d In this work, complex grain microstructure control was achieved for proof-of-concept (POC) parts through a particle size-dependent MPE approach. The relevant tick mark length is 5 mm. A magnified view of the horizontal interface of the matrix (M) and letter (L) regions of the letter 'N' shows a crystal orientation map (IPF-z) obtained by electron backscatter diffraction (EBSD). f The schematic of the letter "N" is superimposed on the outline of the dog bone, and the outward arrow highlights the arrangement of the tensile coupons in the sample. The EBSD crystal orientation and twin boundary maps obtained in the deformed region show a clear response to the deformation in the fine and coarse powder regions. The large-scale EBSD plot of the tensile specimen section is 2 mm scale.
Figure 2: DED-printed monorail melt pool structure and grain morphology of SS316 L.
a Change in grain width and aspect ratio from single-track deposition of process parameter settings 2-18. The error bar represents the ± standard error of the mean. (i) and (ii) IPF-z-plots obtained using parameter set #3 (block fabrication parameters), fine and coarse powder deposition, respectively, on a single-track cross-section. The white dotted line in the diagram demarcates the melt pool boundary. b Schematic diagram of deposited beads and related dimensions and terminology. H, D, and W indicate bead height, melt pool depth, and melt pool width, respectively.
Figure 3: PSD-driven microstructure control in the DED process.
Observation of the melt pool/microstructure of (a) fine, (b) FC and (c) coarse powders. All graduations are 500 microns. d Quantitative measurements of grain width and aspect ratio of fine, FC and coarse samples.
Figure 4: An ML framework for predicting the effect of PSD on the thermophysical properties of a sintered powder bed.
A ML framework illustrates the flow of data from simulating particle microstructures to finite element analysis simulations and ML algorithms. b Thermal conductivity (k) and powder bed density (PBD) response curves obtained by the ML frame. The white arrow indicates the direction in which k increases on the reaction curve. Representative 3D renderings of (c) fine sintered powder beds and (d) coarse sintered powder beds, respectively, analyzed using μ-CT scans, along with their respective simulated particle microstructures. The inset is a FESEM photomicrograph of a fine and coarse sintered powder bed taken at the same magnification.
Figure 5: Applying powder size-driven MPEs to the E-PBF process for convenient large-scale grain control.
IPF-z plots obtained from EBSD of (a) fine powder and (b) coarse powder samples made from E-PBF. The tick marks in (a) and (b) are both 100 μm. c Quantitative analysis of grain size and shape in fine and coarse powder samples. Four different nucleation density No values of 6.64 × 1013, 4.26 × 1014, 3.14 × 1015, and 4.01 × 1015 were determined through experimental observations of the equiaxed grain volume fractions obtained in the E-PBF-fine, E-PBF-coarse, L-DED-fine, and L-DED-Coarse EBSD maps. PDAS stands for Main Dendrite Arm Spacing.
Figure 6: Mechanical response of fine and coarse powder samples made of E-PBF and L-DED.
a Tensile response curves obtained from L-DED and E-PBF samples tested along the build direction. b The engineering stress versus strain (%) obtained from the E-PBF sample highlights the improved mechanical isotropy of the FG equiaxed microstructure obtained from the coarse powder sample. H is horizontal, and V is vertical (or build). c and d IPF-z plots of fine and coarse samples, 3 mm from the fracture surface, respectively. The twin boundaries observed in the EBSD plot are shown in white and are highlighted with dark blue arrows. The insets in e, f) are the corresponding selective electron diffraction (SAED) plots obtained using TEM. c-f) along the plane in the figure.
In conclusion, we not only demonstrated the effect of PSD on the microstructure of AM from the particle feedstock, but also explored its control over the microstructure of specific sites. Both fine and coarse powder final samples retain the original chemical composition of SS316L. We have achieved easy control of the grain microstructure, i.e. a wide range of grain morphology and size, which was previously difficult to achieve. The near-equiaxed FG microstructure of our 3D printed samples is unique and offers extraordinary possibilities for applications such as high strength and ductility, mechanical isotropy and homogeneity, and superplasticity for traditional alloys such as SS316L. Conversely, the near-single crystal microstructure obtained from the fine powder provides a guide for printing nickel-based superalloy single crystals, resulting in ideal high-temperature creep resistance.
Copyright Notice
This article is from the WeChat public account of Materials Science Network
If readers want to understand the current situation and development trend of laser application in new energy vehicles,
Welcome to join us in July 2024 to participate in our Yangtze River Delta G60 Laser Alliance
Conference on the application of laser in new energy vehicles